Download file in PDF format: TRAC 2008: Feeling Like Home: Romanised Rural Landscape from a Gallo-Roman Point of View (pp. 13–24)
Feeling Like Home: Romanised Rural Landscape from a Gallo-Roman Point of View
Introduction The parcelling up of space in Roman provinces, and its political impact, has been well studied from a Roman point of view. The way Romans perceived different kinds of space is quite well documented, from the way grazing lands were considered as underused territory to the use of centuriation as a conqueror’s power display (Purcell 2002). There is little historical evidence, however, that will allow the consideration of the Roman transformation of landscape from a provincial, and furthermore native, point of view in western Gaul. The Late Iron Age way of relating to space and of building place can only be approached through archaeological data on the numerous enclosures. Collecting and comparing these data has allowed the highlighting of the main features of Late La Tène rural settlement layout. In this way, a reconstruction of fragments of the material and cultural landscape of western Gaul has been attempted, from the house outward. From this fragmentary reconstruction of the Late Iron Age sense of place, it appears possible to consider some aspects of rural landscape transformation during the fi rst centuries of this Roman province. As the French ethnologist and prehistorian Leroi-Gourhan wrote, ‘architecture, like dress, is an expression of the identity of its owner and is thus the fi rst to be transformed in a situation of cultural confrontation with a defeater’ (Leroi-Gourhan 1973: 243–246). Two aspects of this cultural process may therefore be questioned, focussing on the layout of rural settlements: which landscape and architectural aspects Gallo-Roman people have adopted because they perceived them as “Roman”, and which cultural patterns were so ingrained in their experience of space that they persisted in the layout of their place.
To consider the transformation of landscape from a provincial point of view, a comparative framework is needed to confront the two datasets, Roman literature and stone buildings on the one hand, and archaeological reports from northwestern Gaul on the other. Two theories concerning the construction of space will be drawn upon, one from psycho-sociology and the other from theoretical architectural corpuses, as they complete each other in their approaches of space, and more interestingly of enclosure. The Roman sense of Place will then be examined within this theoretical framework. Through the analysis of the omnipresent Late Iron Age enclosures and their characteristic features, this paper will attempt to reconstruct the way people interacted with Space and built Place in northwestern Gaul.
How to build a Place from Space: Architecture is not just all about buildings Since the end of the Roman Republic, wealthy citizens were important builders of villas, in the role of dominus aedifi cator, and were directly involved in the building process (Agache 2008). It was so usual to conduct private building enterprises that Otho used this in public as a code for the launching of his conspiracy against the emperor Galba. Tacitus reported accordingly in his Histories that ‘(…) his freedman Onomastus informed him [Otho] that the architect and the contractors were waiting for him. It had been arranged thus to indicate that the soldiers were assembling, and that the preparations of the conspiracy were complete. To those who inquired the reason of his departure, Otho pretended that he was purchasing certain farm-buildings, which from their age he suspected to be unsound, and which had therefore to be fi rst surveyed’ (Tacitus: Histories: I.27). Villa building is a recurring topic in Roman elite correspondence, from Cicero’s reports on several of his own and his brother’s villas, to Pliny Minor’s self-satisfi ed description of different estates. The act of building or embellishing private places thus appears an important part of Roman elite self-promotion. This importance of private architecture is something we understand effortlessly as it seems very similar to the way we consider houses. This feeling of familiarity is quite treacherous as there are some noteworthy differences in the socio-political role of architecture at the time of the Roman Empire. Meanwhile, in Late Iron Age Gaul, this insistence on building is not attested by archaeological data. It seems in fact that buildings are rather undistinguished, and it is usually very diffi cult to identify their purpose. The house is generally identifi ed only by the pattern of refuse scattered around it, and not by its architectural design. It is therefore interesting to try to understand what makes Space a Place theoretically, so as not to focus on our own conception of Place, which is centred on houses and buildings. In his theory on Psycho-sociology of space, Moles explains that building a Place from Space is a two-fold process that implies constructing a valorised “Here” (the architect problematic) and to appropriate this “Here” (the inhabitant problematic) (Moles and Rohmer 1998: 12–13). To differentiate a part of space, it is necessary to create a disruption in the sensorial perception of space, what Moles calls ‘paroi’, that is a wall considered as a partition in general and not as a fence particularly (Moles and Rohmer 1998: 54–60). The strength of ‘paroi’ depends notably on the topological closure provided and the number of perceptions it impaired.
The archetypal ‘paroi’ is in fact the wall, because it interrupts various perceptions, such as view, sound and climate. However, wall in this sense is not only considered as part of a building, but can also be a simple fence. In Fig. 1, Moles outlines how visual fences contribute to the conceptualisation of inside/outside and thus to the appropriation of space. The ‘paroi’ as interruption of a perception can be of various natures, from visual, sound and climatic insulation to more abstract concepts such as juridical or cultural boundaries. Wall and partition are thus fundamental means for the appropriation of space, and it is likely that the fi rst step in appropriating space for an architect would be to build walls. Nevertheless, building walls is not necessary because it is possible to create a sense of Place without any walls or material boundaries, nor is it suffi cient as appropriation is better established in an enclosed zone. Following this concept of Place, the Late Iron Age enclosure and the Roman villa house are two facets of the same way of appropriating space, that is a section of space enclosed by ‘paroi’. Hedges growing on a slope/ditch system and stone-built walls are two different kinds of ‘paroi’ with a similar purpose: creating a sense of Place and of belonging. Immaterial boundaries of a juridical nature are well documented for the Roman Empire, notably in the corpus of land surveyors’ texts. For the Late Iron Age case, we can infer from archaeological data that some kind of immaterial boundaries existed, because a system of enclosures and ditches only covered a small part of the land needed for farming and cattle grazing, but we have no evidence for their nature. It is possible to infer, from what little we know about Late Iron Age society, that they were not juridical, as this kind of appropriation, property in the strict sense, required a certain degree of central administration and the storage of writing documents such as in a land register. The appropriation of Place is thus realised very differently in the Roman Empire and in Late Iron Age northwestern Gaul, but there is ground for comparison thinking in terms of ‘paroi’ and immaterial boundaries. The idea of fence and wall being the primary elements of architecture is quite strong in Norberg-Schulz’s essay titled Intentions in Architecture (Norberg-Schulz 1998). According to him, the role of architecture is to control the interaction between people and environment. A primitive way of controlling the environment is the enclosure of space, creating a frontier between self and outside world. Enclosure is thus a basic form of architecture, which defi nes a domain different from wilderness and gives a visual identity to the community that built it (Norberg-Schulz 1998: 138–140). Buildings, from huts to civic monuments, are a second step in the architectural process of creating Place, as they added an improved environmental control. Norberg-Schulz thus wrote that ‘the fence and the hut are the fi rst expressions of human attempts to master his environment by modifying it according to his own needs and desires’ (Norberg- Schulz 1998: 126). There appear to be two main purposes in planning buildings: one is utilitarian architecture that tries to physically control the environment, the other is monumental architecture that conveys socio-cultural meanings and symbols (Norberg-Schulz 1998: 256–260).
Architecture is, in this sense, the building act that creates an artifi cial climate, a functional environment, a social environment and a symbolic environment, thus defi ning a sense of Place. Late Iron Age enclosures were architectural constructions and can be analysed according to these categories, like any built architecture.
Relying on these theoretical approaches of Space and Place, it is therefore possible to use a common framework to describe the organisation of rural settlements and landscapes for both the Roman period and the Late Iron Age. Using conceptual categories from psycho-sociology and architecture, it will then be possible to identify certain aspects of the evolution that led to the creation of the specific Gallo-Roman villa landscape.
Some aspects of the Roman sense of Space and Place The corpus of Roman documents, varying from literature to epigraphic inscriptions, and from juridical documents to the iconography of wall painting, constitutes an important source in analysing the Roman elite perception of Space and Place. Nevertheless, considering the extensive time span, the evidence enlightening this theme or attitude is sporadic. Therefore, examining Roman concepts of Space and Place in a rural context essentially means looking at some distinctive aspects of Roman elite rural architecture. The importance of the villa as a sociocultural display of fortune and status during the Early Empire has already been emphasised in different works (Gros 2001: 290–291; Agache 2008: 21–23). Some aspects of the architectural design of the villa appeared to be pre-eminent in their Roman depictions and their enduring architectural designs: the legal appropriation of estate as materialised in boundaries and land registers, the visual control of the rural estate from the house, the importance of climatic control in the design of rooms and the social control of the perambulation of visitors.
Figure 1: Visual fence and cognitive appropriation in the creation of inside/outside notions (After Moles and Rohmer 1998: 78, fi g. 10).
The appropriation of Space is a juridical one in the Roman world, where each portion of space under Roman administration was someone’s property, and ultimately the owner was the state itself. To legally possess a place refl ects belief in a superior social force that established this appropriation. This in turn assumes a legal corpus on property and the means to enforce judiciary decisions concerning transgressions of ownership (Moles and Rohmer 1998: 73–74). The Roman Empire used geography to control the newly conquered territories, which were theoretically Roman public property. A land register and census were the administrative tools to control human and fi scal space. Written lists and cartographic documents composed the land register of the centuriated Roman colonies. Such administrative recording of land ownership was not confi ned to provincial territories; maps of estates were drawn by private landlords, even if they were not considered valid juridical testimonies (Nicolet 1988: 213–240). The legal appropriation of estates was quite complex in Roman law, depending on the location – Italy or the provincial territories -, on the personal status of the owner – citizen or not –, but also on the way the property had been transmitted (Gaudemet 2000: 225–234). Property rights were thus dependent on the context. Inheritance rules and practices were also quite different from ours as legacies were not supposed to be transmitted inside the family alone but were used to reinforce clientele relationships. Villa ownership was therefore very variable and did not stay within a specifi c family (Rawson 1991: 204–222). Roman texts indicate that land surveyors insisted on various forms of territory delimitations, encompassing private property. The materialisation of boundaries was very wide-ranging: from stones, trees and other natural demarcations to fences and ditches. Boundary markers and demarcations were considered sacred and meddling with them was a very grave offence that was sanctioned severely (Chouquer and Favory 2001: 181–192). Property was thus diverse in nature although protected by laws and religious beliefs.
Boundary markers demarcated the limits of the rural estate but the appropriation of the villa territory was also created by a visual control of the rural estate from the residence. Italian villa layouts were hence designed to display the attractiveness of the views offered from the house on the surrounding landscape. Vitruvius, stating the difference between urban and rural houses, wrote that their architecture was similar except for the fact that the latter had a concentric design: ‘(…) in town atriums are usually next to the front door, while in country seats peristyles come fi rst, and then atriums surrounded by paved colonnades opening upon palaestrae and walks’ (Vitruvius: On Architecture: VI, 5, 3; Gros 2001: 267). Italian rural residences that have been well surveyed, such as Settefi nestre in Etruria, are organised in this way, with a view of the countryside from different rooms of the villa house. The Settefi nestre villa was set on the peak of a slope, with the main storey built on cryptoporticuses, so that the buildings dominated the surrounding hilly countryside. Doors and apertures opened towards fi elds, a terraced garden and the hill at the back. In Campania, the villa of Madonna Grande overlooked a valley that led to the Falerna plain and farther away to the sea. For Vallat, this villa controlled the surrounding space, and its economic takeover of the surrounding ground is materialised by the ideological command of a heightened site, the elevation being artifi cially accentuated with two terraces and a cryptoporticus, with the house being built upon the cryptoporticus (Vallat 1983). On the bottom of the Monte Massico slope, about ten villas present a shared design, their architecture relying on the cryptoporticus terracing, even when it was not required by the relief of the land. Italian villa architecture was designed to provide a view of the surrounding landscape. The idea of visual control over the whole domain, through a vista of people farming the land, livestock grazing pasture or of different kinds of fi elds, is quite developed in Roman literature. In Cicero’s and Pliny Minor’s letters, lavish depictions of the different landscape views offered from their villa residences can be found. Pliny Minor thus described his villa in Tuscany as a wide natural amphitheatre, and continued with the depiction of each and every kind of land production possible in this domain (Pliny Minor: Letters: V.6). The focus was on the variety of agricultural production: vineyard, pasture, people at work… This was an important characteristic of a domain for the Roman elite. Ancient literature thus celebrated the villa “landscape of production” (Purcell 1995). In this way the Roman villa, a productive estate bringing wealth to its owner, was architecturally staged to let him enjoy an agro-pastoral show of day-to-day rural labours. The location of the residence in the estate is accordingly chosen to offer varied vistas on the surrounding landscape.
This villa architecture linked the house to the surrounding world despite creating a microcosm. Just as enclosure defined a different domain from wilderness in Norberg-Schulz’s essay, the Roman villa was conceived as an island of controlled Nature. The layout of the rooms was therefore designed to offer winter and summer suites, by playing with the cardinal orientation of the rooms, the careful design of the openings, a cryptoporticus for cooling down and a hypocaust for heating. In Pliny Minor’s letters, this climatic control emerged from various descriptions of his villas and of their architectural designs. Columella portrayed accordingly the ideal layout of the villa house: ‘the landlord’s house will be divided into summer and winter suites. The winter suite bedrooms will be exposed to the winter rising sun and the dining rooms to the equinox setting sun. The summer suite bedrooms will be exposed to the equinox midi and the dining rooms to the winter rising sun. The baths will be turned toward the summer setting sun, so they will be well lighted from midday to evening. The covered walks will be under the equinox midi so they will receive a great deal of sun in winter and little in summer’ (Columella: On Agriculture: I. VI.). The length and precision of the description of the layout of the rooms shows the importance of climatic control in the living rooms of the landlord’s house. The ideal of otium carried by Roman villas imprinted thus their design as well as the Roman ideal of controlled nature.
The Roman sense of Place was built on an idea of property that was juridical and sacred at the same time, materialised in boundary markers that demarcated the estate. The location and architecture of the villa was chosen to offer visual control over a productive landscape, but at the same time offered a controlled climatic environment for leisure time. Other characteristics of Roman villa design have been identifi ed, such as the social control of visitor access depending on the visitor’s status, but these are more evident in urban house architecture or in Late Empire villas. The three facets of rural estate appropriation described above – juridical boundary, visual and climatic controls – are best documented in Roman Empire texts.
The architecture of Late Iron Age enclosures in northwestern Gaul Late Iron Age rural settlements in northwestern Gaul are characterised in archaeological surveys by their ditched enclosures, which is their main remaining feature. The ditches are only the last relics of the actual limits of the settlement, and their main function may have been to provide the building materials necessary for the erection of an embankment. Moreover, archaeobotanical data, when available, indicate the presence of a planted hedgerow composed of different trees, shrubs and other plants. It would have created a fence several metres high, of some signifi cance in the landscape. Caesar even mistook these quickset hedges for little patches of woodland, writing in his account of the Gallic wars that woods generally surrounded Gallic homes. Creating Place is generally achieved by partitioning space because, as Moles wrote, ‘the notion of partition is integral to the idea of spatial appropriation. Human beings only conquer space by dividing it, organising it and reducing it to their own scale, and by making real its subdivisions’ (Moles and Rohmer 1998: 62–63). In northwestern Gaul, enclosures surrounding the settlement functioned as partitions, creating a signifi cant visual barrier that demarcated the settlement. Enclosures were, in this sense, a vegetal architecture expressing the way their builders related to place and territory.
Although Late Iron Age enclosures created a sense of Place and a feeling of belonging, they appear to have defi ned housing and domestic space but not the entire rural estate. The hedge and embankment system does not appear to have been used to delineate the borders of farmland, and there are few other archaeological remains that can be related to boundary markers. Even where other enclosures or ditches demarcated some parcels around the farmstead core, probably for farming purposes, they lack the visual impact of the farmstead’s hedgerows, with only little evidence of sparse trees and no shrubs, as well as shallower and narrower ditches implying smaller banks. These structures are generally not very extended, creating only a few parcels in the immediate vicinity of the farmhouse and not covering enough land to sustain a household. These small fi eld systems are usually less than a few hundred metres in length. Thus, farmstead enclosures seem not only to have characterised the place of residence, but also to have conveyed the appropriation of the surrounding space by the inhabitants (Courbot-Dewerdt 2005).
Inside the vast space defi ned by the enclosure, the settlement was composed of several disseminated timber and daub buildings. Among them, the house appeared quite undistinguished, with an architecture that was very similar to that of the other farm buildings. Its identifi cation is generally based on the refuse discovered in the nearby ditch or surrounding pits. The analysis of the internal layout of enclosed settlements in northwestern Gaul uncovers no strict building plans. However, the house was generally set at the farthest side from the main enclosure entrance, and was sometimes even separated from the rest of the farm by partitions. In some cases, however, partitions of the interior space defi ned several areas, one of them generally dedicated to the main building. In Fig. 2, the Late Iron Age house appears to have been situated in the innermost place, opposite the principal enclosure entrance. Enclosures and ditched partitions organised therefore a progress from outside world to private space. This fenced architecture structured a controlled access from outer space to domestic and then private areas. A main path in the enclosure led to a working area with granaries, kilns and other farm structures. From this working space, a path led to the residence, sometimes through a specifi c passageway when the house was situated in another enclosed space. In some cases, a footbridge over the enclosure ditch has even been discovered. The progression from the outside world into the privacy of the house appears therefore to have been organised to control access. The architecture of enclosure puts some emphasis on the entrance. While most northwestern Late Iron Age enclosures were rectangular in design, some of them appeared to have a trapezoidal layout with a longer entrance side, so that someone going inside would have thought the enclosed area larger than it really Figure 2: Several examples of Late Iron Age rural settlement layout in northwestern Gaul). was. In other cases, the ditches were wider and deeper near the entrance, implying a higher and thus more impressive embankment. Doubling the front side of the enclosure with a parallel ditch achieved the same purpose. Walking through such an impressive passageway probably reinforced the feeling of leaving the outside world to access a private place.
This architecture of enclosure was also used to emphasise the evolution of the settlement through time. Archaeological records demonstrate that enlargement of the initial enclosure was a common feature during the Late Iron Age. Examples of these enlargements illustrate the symbolic importance of this rebuilding of the fence (Fig. 3). Actually, the layout of the enclosure did not change, and no new spatial organisation was realised. The new surrounding fence did not enlarge the inner space either, and there is no real evidence that the fi rst enclosure was actually removed at this time. The cyclic management of trees can partly explain this double row of fences, one line of old trees ready to be used and a new, growing one, but this was not necessary. This architectural evolution should also have enhanced the visual and monumental aspect of enclosure in the surrounding landscape. The Late Iron Age enclosure created a visual focal point in the landscape, expressing the appropriation of the farmland, and its enlargement was probably more due to this expression of Place than to practical reasons, considering the amount of work involved.
The layout and fence management of the enclosure appear to have created a strong feeling of belonging to the place. Many architectural functions of the Roman villa residence, such as access control, visual control and appropriation, were in fact fulfi lled during the Late Iron Age by the enclosure. The undistinguished appearance of the residence reinforced the impression that the real architecture was the hedge and embankment enclosure in Late Iron Age northwestern Gaul. The appropriation of farmland was henceforth sustained by the visual focal point created by the enclosure, standing out in the rural landscape, while the Roman sense of Place was based on a view of a productive villa landscape.
The building of the Gallo-Roman villa Roman architecture at the beginning of the Empire was already the result of acculturation processes, a specifi c mix of Central Italian and Greek symbols, forms and types. The integration of northwestern Gaul in the Roman Empire has led to another acculturation with the architectural evolution of the Late Iron Age rural settlement that culminated with the creation of the Gallo- Roman villa. The Late Iron Age conception of place, with its vegetal architecture of the enclosure and the wood and daub building techniques, was quite alien to the Romans. Caesar’s and Vitrivius’ reactions to Gallic architecture illustrate the cultural distances: the fi rst did not understand the enclosure’s purpose and gave a unsatisfactory climatic explanation of the phenomenon and the other had compared the wood and daub constructions to swallows’ nests and categorised them as “primitive architecture”. The fact that Roman authors like Tacitus acknowledged Gallo- Roman rural settlements as villas assumes some important evolution in their architecture and the assimilation of different cultural concepts.
After Gaul had been defeated by Rome, its territory became the legal property of its conqueror. If there were immediate changes in land possession in the wake of the Gallic wars, the main change therefore was that legal appropriation was imposed, ruled by Roman laws. Without important implantation of Roman colonies and thus land expropriation and subsequent centuriation in northwestern Gaul, the transformation was apparently not dramatic. Changes in landscape organisation can, however, be observed. During the Late Iron Age, the physical demarcation of farmlands appears to have been limited to the immediate surroundings of the settlement. From the beginning of the Early Empire onwards, a road network became more visible in the landscape and it appeared more ordered. In the meantime, a larger-scale parcelling up of land occurred, structured by the new road network. Long-distance ditches divided the territory, generally in mostly rectangular pieces. The appropriation of farmland was therefore transformed in its material boundaries as in its underlying conceptions.
The fi rst noticeable changes in the definition of Place occurred in the architecture of the residence. The slightly blurred shape of Late Iron Age thatched constructions was progressively replaced by rectangular buildings with tiled roofs. The significant change in architecture was not so much the use of new building techniques but the adaptation of earlier construction practices to accommodate a new aesthetic: a regular and geometric building with whitened walls and a tiled roof. At the same time, the internal layout was profoundly changed with the appearance of a porticus giving access to a single row of rooms (three in most early villa houses). In his essay on Roman builders: a study in architectural process, Taylor has listed some of the main Roman architectural features: ‘the arch, the vault, the Greek or Tuscan columnar order, the pediment, the bathing suite, the garden courtyard, the ornamental fountain, the aqueduct (…)’ (Taylor 2003: 7–9). Some of these typically Roman architectural elements, like the porticus and pavilion layout or the bathing suite, were gradually adopted in Gallo-Roman villas, while others are very rarely documented in archaeology, such as the garden courtyard or ornamental fountain. Following the successive transformation of the residence through adopting a Roman aesthetic, the organisation of the settlement underwent profound changes. The tiled roof and whitened walls had given a very Roman flavour to the first century A.D. rural residences that hence stood out among the thatched farm buildings.
Figure 3: Several examples of the architectural evolution of Late Iron Age enclosures in northwestern Gaul.
The location of the house evolved from a Late Iron Age design, as it was set at one side of the enclosure, opposite the main entrance. The enclosure remained a signifi cant feature of the architecture of Gallo-Roman rural settlements, even though the hedge and ditch system was progressively replaced by fences and then by masonry walls. The integration of an Iron Age pattern in what were basically adopted foreign building techniques shows the strength and signifi cance of the enclosure in La Tène society (Courbot-Dewerdt 2005). The residence was separated from farm buildings by a wall or fence. The other farm constructions were generally lined up in two rows in a long rectangular courtyard, overlooked by the residence. The progression from outside to private space was thus emphasised by the layout of the enclosure. This Late Iron Age spatial design, and its enhancement during the Gallo-Roman period with enlarged courtyards, meant that any visitor would have had to walk through the whole farm courtyard, getting, in the process, a good view of storage buildings, stables and the like, before reaching the residence proper. This design implies that the vista from the residence would have been very different from that in Roman villas. The fenced courtyard and the setting of the house facing the inside of the enclosure would not have allowed any view on the surrounding countryside and the “landscape of production” so dear to the Roman elite. The design of the large Gallo-Roman courtyards, with rows of buildings on both sides of the residence, even if having roots in the Late La Tène conception of space, seems to stage the farm buildings and thus the estate’s productivity. The main sight from the residence was the courtyard itself, probably as busy with men at work and livestock as the Roman villa’s surrounding lands. The Gallo-Roman villa design seemed thus to have expressed in another way the same feelings conveyed by the “landscape of production” of the Roman elite.
In northwestern Gaul, rural settlements were marked by their Late La Tène architectural heritage but changes are noticeable in spatial conception. They can be traced back to a more Roman conception of space: the signifi cance of the residential architecture and the importance of staging productivity. In this respect, Gallo-Roman rural settlements can be called villas as they endorsed in their own way some Roman cultural concepts of the villa.
The integration of northwestern Gaul as a province of the Roman Empire has led Gallo- Romans to progressively change their way of interacting with space. Analysing in turn Roman and Gallic concepts of Space and Place in a common theoretical framework leads to a better understanding of the similarities and differences in their cultural management of space, and thus to the identifi cation of continuities and discrepancies in the Gallo-Roman building of Place. The large-scale parcelling up of farmland with roads and ditches materialised the introduction of a new way of relating to territory, based on the appropriation of space in a Roman legal context. Transformations in house architecture were multifold, showing the degree of socio-cultural changes: a new importance given to the house, the enlargement and multiplication of rooms, and the adoption of some Roman architectural elements like the porticos or tiles. Meanwhile, the layout of Late Iron Age rural settlements is emphasised in new materials and a new organisation of farm buildings. Contrary to the Roman experience of place, which focussed on the rural estate and the visual control of its surrounding productive landscape, the Gallo-Roman enclosure and its two-courtyard system staged in a new way the productivity of the farm. The spatial progression from outside world to private house invited the Gallo-Roman visitor to cross the entire length of the rural settlement, appreciating thus along the way the various storage buildings as well as people at work or cattle grazing. The role of the enclosure in defi ning place appears thus to be reframed in new cultural material, with new architectural techniques and a new notion of place. In this sense, the Gallo-Roman approach appears to have been an original way of experiencing Space and Place in the Roman cultural framework.
Centre de recherche HALMA-IPEL – UMR 8164 (Université de Lille III)